The Study Of SMEs Credit Scoring Model Based On Support Vector Machine

SMEs (Small and Middle Sized Enterprises) have developed into an important part of the national economy, since the reform and open policy. It has made historical contributions in many aspects, such as scientific and technological innovation and progress, creating products, increasing employment, economic prosperity and so on. The difficult financing problem has deeply limited the development of SMEs in China. As the main source of financing for SMEs, commercial banks have the responsibility and obligation to improve the financing environment for SMEs and to better serve the real economy.Compared with traditional lending techniques, credit scoring realizes handling loans automatically to a large degree and thus improves the efficiency of approving and screening loan applicants, which helps the bank price these loans more exactly based on credit risk involved. Large quantities of studies indicate that credit scoring works well in small business lending and can improve their credit availability. Meanwhile, it also has many other advantages, such as reducing the transactions costs, controlling and preventing the financial risks, improving the lending rate of SMEs.In this paper, the SMEs credit scoring model is explored based on the SVM (support vector machine) method. Firstly, based on reviewing of related literatures and research results, relating concepts are defined, and the applicability of credit scoring method to solve the financing difficulties is analyzed. Secondly, the Support vector machine theory and building process of support vector machine’s credit scoring model are introduced. Thirdly, based on SVM method, credit scoring model is built. It is indicated that compared with Neural network meta-model, support vector machine has a stronger classification ability to treat the problem of credit scoring. The classification accuracy is relative with the selection of support vector machine’s kernel function, and the classification result of the radial basis kernel function is of the best effect. Finally, both the application of credit scoring models and corresponding recommendations are proposed.